Exposure of cells to ionizing radiation (X-rays and gamma-rays) results in the formation of double-stranded breaks (DSBs) in chromosomal DNA, a particularly deleterious form of genomic injury. In all organisms, one of the main ways to repair DSBs is by homologous recombination using undamaged DNA present elsewhere in the cell. The purpose of the experiments in this proposal is to analyze the enzymology of this reaction in the yeast Saccharomyces cerevisiae. Specific Aim I: One of our goals is to identify and understand proteins that assemble at DSBs and catalyze early steps in their repair. To address this issue, we have been utilizing a powerful model system for recombinational repair: mating-type switching. We have identified a new protein, called YZ binding protein, which is required for DSB formation in this system. In the first Specific Aim, we describe experiments to analyze the function of this protein in detail. These studies will allow us to progress to the identification and characterization of additional factors involved in early steps of DSB repair. Specific Aim II: Another of our goals is to understand proteins that generate heteroduplex DNA (hDNA), the central intermediate in recombinational repair. Using an in vitro D-loop assay, we have purified a large complex from yeast extracts that catalyzes hDNA formation in an ATP- and homology-dependent reaction. Optimal product formation by this complex requires functional Rad51p and Rad52p, each of which is known to play an essential role in recombinational repair. In the second Specific Aim, we describe experiments to further characterize this activity. Analysis of such a "recombination machine" will dramatically improve our understanding of the enzymology of recombinational repair. Significance: DSBs are extremely dangerous to the cell. They can, for instance, induce cell death or genomic rearrangements if left unrepaired. DSBs also have positive roles in cells, i.e., they initiate the meiotic recombination reaction required for the proper segregation of chromosomes into gametes. Finally, it has been demonstrated that human Rad51p, a central protein in recombinational repair, is associated in vivo with the BRCA1 and BRCA2 genes, suggesting that breast cancer results from a defect in recombinational repair. The analysis of recombinational-repair proteins in simpler and more tractable organisms, like yeast, will therefore provide invaluable insights into several fundamental and important cellular processes.